The recent report of an extensively drug resistant (XDR) tuberculosis (TB) outbreak in Kwazulu-Natal, South Africa has alerted health authorities to the worsening global TB epidemic. TB control is increasingly compromised by the rise in drug resistant, multidrug resistant (MDR) and XDR-TB strains, reported in most countries surveyed. The goal of this research proposal is to explore host and pathogen factors that contribute to the failure of treatment of MDR-TB, ultimately leading to the emergence of XDR-TB strains. These studies will be done in South Africa, a country with some of the highest rates of TB worldwide. We will carry out two cross-sectional population-based studies of newly diagnosed MDR-TB patients from Gauteng Province, South Africa to characterize the nature and extent of the M/XDR-TB epidemic, to examine the diversity of MTB strains and drug resistance mutations represented in this population and to evaluate trends in the epidemic over the 5-year project period. We will also investigate the association between resistance to drugs in addition to INH and RIF and time to sputum culture clearance in patients undergoing treatment for MDR-TB. For these studies, two patient cohorts will be defined: Fast Responders (sputum culture clearance by the median time) and Slow Responders (failure to achieve sputum culture clearance by the median time). We will separately evaluate the impact of drug resistance (phenotypic/genotypic) that is pre- existing in the infecting bacillary population and drug resistance that is acquired de novo during treatment. Finally, we will explore selected nutritional and immunologic characteristics associated with poor response to treatment in the same patients undergoing treatment for MDR-TB. We will examine whether the time to sputum culture clearance and/or mortality observed among MDR-TB patients are associated with specific nutritional and/or immunologic profiles. The information gained from the cross-sectional studies can help to evaluate the extent to which XDR-TB may be attributed to primary transmission or acquired resistance and whether any M/XDR-TB strains are emerging in the region. The results of our clinical studies will help to elucidate factors that can contribute to poor response to treatment in MDR-TB patients, thereby fueling the increased incidence of XDR-TB. These data can inform current and future TB control strategies. As every additional M/XDR-TB patient is a public health hazard and a burden to the TB control program, our ability to identify factors that drive the epidemic will allow preemptive and targeted interventions. These may include improved treatment regimens, better approaches for monitoring patients during treatment, more intensive follow-up and other control strategies aimed at reducing the emergence of XDR-TB.